Question Self Built PC Dead After 3 Years

FateCPS

Commendable
Jun 29, 2019
11
0
1,510
0
I should start off with my specs, Ryzen 2700x, Msi Gaming Gtx 1080, Asus X470-F Mobo, and TridentZ 3200 Ram. I built this PC back in late 2017 /early 2018 and since resolving my 1st problem of a DOA Moba I've had no issues. Today while watching youtube I went to pause the video and once I did, the PC froze - So naturally, I figured just a restart would fix it. I restarted the computer and it power cycled 4-5 times, lights would come on for a split second, then quickly turned back off before repeating (I believe the Mobo light was red). On the 5th cycle, the computer stayed on with a yellow/orange Mobo light. I'm not getting a display connection on either monitor so I'm not sure how to proceed with testing. The LEDs and fans appear to run correctly except for the GPU fans (Which typically dont turn on until later anyways I believe?).

If you have any idea on how I should proceed I'd really appreciate it, I do have a 2nd pc but it's ancient so the only swap I could tryis the GPU if you think that would help

Edit: Just remembered that I had installed two extra sticks of ram a few months back. Theyre the same model but I dont remember if I did the bios <Mod Edit> to set their actual speed like I did with the 1st two. Could that be it? But why would it only happen months later
 
Last edited by a moderator:
The Yellow/Orange led that stays lit is the troubleshooting led for DRAM.
Post will not continue until the problem is fixed hence no display or output.

You should check your RAM modules by testing with a single module in the slot meant for single channel mode and try switching modules around. This way you can check for a failing module or MB slot.
You should also test all of your DIMMs using Memtest86 and booting from the USB. Run four passes and should there be any errors then you have a fail. (RMA the full kit if any errors)
You can then run memtest on all modules individually to determine the culprit.

You installed extra RAM and it appears you mixed kits which is a NO NO.
You may have a mismatch as RAM kits with exactly the same specs can mismatch due to slight variations in Latency/Density. DIMMs are binned at the factory to ensure this problem does not happen.

If this proves to be a mismatch then return the recent kit you installed and get a single kit the size and frequency that your CPU and MB support.

When you are able to access Bios then update to latest Bios version and latest chipset driver.
 
Reactions: FateCPS and Why_Me

FateCPS

Commendable
Jun 29, 2019
11
0
1,510
0
The Yellow/Orange led that stays lit is the troubleshooting led for DRAM.
Post will not continue until the problem is fixed hence no display or output.

You should check your RAM modules by testing with a single module in the slot meant for single channel mode and try switching modules around. This way you can check for a failing module or MB slot.
You should also test all of your DIMMs using Memtest86 and booting from the USB. Run four passes and should there be any errors then you have a fail. (RMA the full kit if any errors)
You can then run memtest on all modules individually to determine the culprit.

You installed extra RAM and it appears you mixed kits which is a NO NO.
You may have a mismatch as RAM kits with exactly the same specs can mismatch due to slight variations in Latency/Density. DIMMs are binned at the factory to ensure this problem does not happen.

If this proves to be a mismatch then return the recent kit you installed and get a single kit the size and frequency that your CPU and MB support.

When you are able to access Bios then update to latest Bios version and latest chipset driver.
Ill give this a try tomorrow when I get the chance thank you
 

FateCPS

Commendable
Jun 29, 2019
11
0
1,510
0
The Yellow/Orange led that stays lit is the troubleshooting led for DRAM.
Post will not continue until the problem is fixed hence no display or output.

You should check your RAM modules by testing with a single module in the slot meant for single channel mode and try switching modules around. This way you can check for a failing module or MB slot.
You should also test all of your DIMMs using Memtest86 and booting from the USB. Run four passes and should there be any errors then you have a fail. (RMA the full kit if any errors)
You can then run memtest on all modules individually to determine the culprit.

You installed extra RAM and it appears you mixed kits which is a NO NO.
You may have a mismatch as RAM kits with exactly the same specs can mismatch due to slight variations in Latency/Density. DIMMs are binned at the factory to ensure this problem does not happen.

If this proves to be a mismatch then return the recent kit you installed and get a single kit the size and frequency that your CPU and MB support.

When you are able to access Bios then update to latest Bios version and latest chipset driver.
I was able to pass memtest with my original 2 sticks, the other two still wouldn't boot up. I was under the impression that mismatched sets would only effect effective performance and may not be as optimal as a full set. Mixing the two wasn't the cause of them breaking was it? Im guessing they were shipped with a defect but I don't know much about the subject
 

InvalidError

Titan
Moderator
When you mix kits, they may or may not work together. You tested your original memory with memtest86, you should try repeating the test with only the new RAM installed.

If both the old and new memory pass tests separately but fail when combined, the problem you may be running into is that memory IO in the CPU and/or DIMMs may have slowed down over time kind of like tolerances in new engines loosen up over the break-in period and cannot keep up with your old speed + timings. You may be able to get them working together again by budging your memory speed down one peg.
 

FateCPS

Commendable
Jun 29, 2019
11
0
1,510
0
When you mix kits, they may or may not work together. You tested your original memory with memtest86, you should try repeating the test with only the new RAM installed.

If both the old and new memory pass tests separately but fail when combined, the problem you may be running into is that memory IO in the CPU and/or DIMMs may have slowed down over time kind of like tolerances in new engines loosen up over the break-in period and cannot keep up with your old speed + timings. You may be able to get them working together again by budging your memory speed down one peg.
That makes sense, however the 2nd pair of sticks won't boot at all. They power cycle until they settle with the orange light error on the Mobo
 

InvalidError

Titan
Moderator
Sent in the RMA request today. Do you think they were damaged from the start? Or from being mixed with the other set? Ive never heard of that happening
For the most part, ICs either work or they don't. Sometimes defects will make it through QC undetected, maybe your 2nd set got a material or manufacturing defect that took a while to manifest.
 

animekenji

Distinguished
Dec 31, 2010
196
32
18,690
18
It's funny how the longevity on electronics is all over the place. I have machines here going all the way back to XT class PC's, Atari's Commodores, Apple II's and others that are still working.
 

InvalidError

Titan
Moderator
It's funny how the longevity on electronics is all over the place. I have machines here going all the way back to XT class PC's, Atari's Commodores, Apple II's and others that are still working.
When electronics are going to fail from a manufacturing or material defect, the bulk of them will do so within the first 3-4 years, except when it is engineered failure in the form of under-rated caps picked to fail just after the warranty expires. I had to bust the AC adapter brick for one of my external HDDs a couple of weeks ago because its input filter cap had died, preventing it from outputting continuous output voltage.

Today, my desktop USB hub caused a system-wide USB driver chain reset while I was copying files from an external HDD to an SD card and it turned out its 5V 3.5A adapter couldn't sustain more than 500mA before dropping below 4.75V anymore. Haven't cracked it open yet since I happened to have a spare from a dead USB2 hub that can still put out 4A. The USB2 hub was around 15 years old at the time of failure. Couldn't find anything wrong with supply voltages or anything else I could measure inside the hub, so my best guess is that it may simply have been the SPI-EEPROM getting corrupted by leakage: 15 years is just about the soonest this can be expected to become an issue and those early-gen USB2 hubs were running kind of hot, which is bad for data retention.
 

ASK THE COMMUNITY

TRENDING THREADS